Currently, I've indexed many documents in ElasticSearch and when I did a search on max score, the number exceeds 1.
How can I obtain a similarity search, which is between 0 and 1 ?
Should I configure ElasticSearch better ? or I should change my mapping?
Thanks in advance for your advices.
Lucene scoring is not classical cosinus-similarity function based on tif id vectors, but some modification wich takes into account other factors like boost on field or query terms. You can read a mode details about scoring formula at page. So the simplest way to get scoring between a 0.0 and 1.0, just norm score on max_score from elasticsearch response (it coincides with score of the first document).
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With